[Bluez-devel] DBUS service support - Initial library

Claudio Takahasi

author

2005-04-28 20:00 UTC #0 lore raw

Hi Marcel,

I wrote the skeleton of the library. Please verify if I=20 am following the BlueZ principles.=20
At the moment, the code is not fully integrated=20 with BlueZ daemons and applications. Some functions=20 shall be changed due function outputs: the result=20 should be returned instead of displayed on the=20 screen.

It is only a initial draft. The idea is provide a=20 library to user mode application controls bluetooth=20 connections. Please validate my suggestions before=20 I start coding and verify if this library can be=20 useful for BlueZ community. I know that some features of this library are provided by "pand", but I need control connections from a user mode application=20 or DBUS application.

The code is available in the link below:
http://www.cin.ufpe.br/~ckt/panz_lib.tar.gz

The following changes were done:
1. Makefile.am
added libpanz in the SUBDIR
2. configure.in
added libpanz/Makefile in the AC_OUTPUT
3. Added libpanz directory
the files are in libpanz.tar.gz file
4. Common hcitool functions
hcitool functions used by the library were moved to hci_common.c
5. Dynamic Link Local Layer addresses
The automatic IP assignment will be based on the=20 following IETF draft=20
http://files.zeroconf.org/draft-ietf-zeroconf-ipv4-linklocal.txt

There are a lot of chalenges until I reach=20
my objective: integrate with a DBUS application.

Best regards,
Claudio.

-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-04-29 09:47 UTC #1 lore raw

Hi Claudio,

Quoted 8 lines

I wrote the skeleton of the library. Please verify if I am following the BlueZ principles.
At the moment, the code is not fully integrated with BlueZ daemons and applications. Some functions shall be changed due function outputs: the result should be returned instead of displayed on the screen.

Quoted 8 lines

It is only a initial draft. The idea is provide a library to user mode application controls bluetooth connections. Please validate my suggestions before I start coding and verify if this library can be useful for BlueZ community. I know that some features of this library are provided by "pand", but I need control connections from a user mode application or DBUS application.

don't move anything around the utils directories. It is easier to do that later on when it is clear what to do. If you need some stuff that is not in the Bluetooth library duplicate the code for now.

Since this is all about pand. Extend it with your functionality and put everything in the pand directory. I don't see any need for another library especially when you start talking about D-Bus integration.

Regards

Marcel

-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-04-29 12:26 UTC #2 lore raw

Hi Marcel,

Let me try justify why I need a new library.
1. A DBUS app should not contain specific bluetooth/Wi-Fi/GPRS connection c= ode.
In my opinion, it is better abstract the connection details.
2. Features required by DBUS app: signal strength, connection establishment= ,=20
disconnection, new device notification(search for service
frequently), automatic
IP assigment(with IP conflict resolution and IP announment),
device information,
active connections, security info, connection lost, ...
All these features should be provided by a BlueZ library. At the
moment, pand
and hcitool provide part of these. However, they are command line
application
and a DBUS app should use a workaround approach: "execv" function
3. Consider an embedded device, control connections using a DBUS app=20 is the better way to this task. Execute a command line app is not
appropriated
to handle this context. An embedded app using wireless connection should abstract connection establishment/control details.

This is the scenario that I am working on, if I can't contribute developing this library,
I will try move all these required features to DBUS app level without add functions/library
to BlueZ.

What is your opinion about this?

Best regards,
Claudio.

Quoted 32 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 8 lines
I wrote the skeleton of the library. Please verify if I am following the BlueZ principles.
At the moment, the code is not fully integrated with BlueZ daemons and applications. Some functions shall be changed due function outputs: the result should be returned instead of displayed on the screen.
Quoted 8 lines
It is only a initial draft. The idea is provide a library to user mode application controls bluetooth connections. Please validate my suggestions before I start coding and verify if this library can be useful for BlueZ community. I know that some features of this library are provided by "pand", but I need control connections from a user mode application or DBUS application.
=20
don't move anything around the utils directories. It is easier to do that later on when it is clear what to do. If you need some stuff that is not in the Bluetooth library duplicate the code for now. =20
Since this is all about pand. Extend it with your functionality and put everything in the pand directory. I don't see any need for another library especially when you start talking about D-Bus integration. =20
Regards
=20
Marcel
=20
-------------------------------------------------------

Quoted 9 lines

SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=3D105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-04-29 12:43 UTC #3 lore raw

Hi Claudio,

Quoted 20 lines

Let me try justify why I need a new library.
1. A DBUS app should not contain specific bluetooth/Wi-Fi/GPRS connection code. In my opinion, it is better abstract the connection details.
2. Features required by DBUS app: signal strength, connection establishment, disconnection, new device notification(search for service
frequently), automatic
IP assigment(with IP conflict resolution and IP announment),
device information,
active connections, security info, connection lost, ...
All these features should be provided by a BlueZ library. At the
moment, pand
and hcitool provide part of these. However, they are command line
application
and a DBUS app should use a workaround approach: "execv" function
3. Consider an embedded device, control connections using a DBUS app is the better way to this task. Execute a command line app is not
appropriated
to handle this context. An embedded app using wireless connection should abstract connection establishment/control details.

Quoted 5 lines

This is the scenario that I am working on, if I can't contribute developing this library,
I will try move all these required features to DBUS app level without add functions/library
to BlueZ.

think about adding all you need to pand and then start it for example with a --dbus option to activate that service via D-Bus.

When writing code use the BlueZ coding style. This means, use tabs. For the patches, create them with "cvs diff -u" from the BlueZ CVS and make sure that you executed "make maintainer-clean" first.

Regards

Marcel

-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-04-29 22:20 UTC #4 lore raw

Hi Marcel,

My application will provide a DBUS service and APIs to alows user mode applications interact with diferent types of Wireless/Wired connections. This is the reason to create a BlueZ library based on pand and hcitool functions. My application should the able to retrieve bluetooth device information, including establish connection, disconnect, link status, new devices/services, ...
The DBUS service will not be provided at BlueZ level.

The current DBUS message in the hcid is used to handle security/authentication issues. This message has a different proposal. Therefore, use the same macro "--enable-dbus" to enable/disable is not a good ideia.

Another important information is: the user mode application doesn't need root privileges to create a bluetooth connection, because it will request using APIs and DBUS service.

Can I create a library in the pand directory and reuse the common codes?
I suggest a directive like "--enable-panz" because this library aims provide functions to hide connection details.
Do you agree?

Regards,
Claudio.

Quoted 4 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 2 lines
Let me try justify why I need a new library.
1. A DBUS app should not contain specific bluetooth/Wi-Fi/GPRS connecti=

on code.

Quoted 2 lines

Quoted 2 lines
In my opinion, it is better abstract the connection details.
2. Features required by DBUS app: signal strength, connection establish=

ment,

Quoted 35 lines

Quoted 16 lines
disconnection, new device notification(search for service
frequently), automatic
IP assigment(with IP conflict resolution and IP announment),
device information,
active connections, security info, connection lost, ... All these features should be provided by a BlueZ library. At the
moment, pand
and hcitool provide part of these. However, they are command line
application
and a DBUS app should use a workaround approach: "execv" function
3. Consider an embedded device, control connections using a DBUS app is the better way to this task. Execute a command line app is not
appropriated
to handle this context. An embedded app using wireless connection should abstract connection establishment/control details.
Quoted 5 lines
This is the scenario that I am working on, if I can't contribute developing this library,
I will try move all these required features to DBUS app level without add functions/library
to BlueZ.
=20
think about adding all you need to pand and then start it for example with a --dbus option to activate that service via D-Bus.
=20
When writing code use the BlueZ coding style. This means, use tabs. For the patches, create them with "cvs diff -u" from the BlueZ CVS and make sure that you executed "make maintainer-clean" first.
=20
Regards
=20
Marcel
=20
-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans!

Quoted 8 lines

Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=3D105hix
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
=20

--=20
_______________________________________
Claudio Takahasi
INdT - Nokia Technology Institute
Phone: +55 81 21223034
Recife - Pernambuco - Brazil

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-04-30 00:04 UTC #5 lore raw

Hi Claudio,

Quoted 20 lines

My application will provide a DBUS service and APIs to alows user mode applications interact with diferent types of Wireless/Wired connections. This is the reason to create a BlueZ library based on pand and hcitool functions. My application should the able to retrieve bluetooth device information, including establish connection, disconnect, link status, new devices/services, ...
The DBUS service will not be provided at BlueZ level.
The current DBUS message in the hcid is used to handle security/authentication issues. This message has a different proposal. Therefore, use the same macro "--enable-dbus" to enable/disable is not a good ideia.
Another important information is: the user mode application doesn't need root privileges to create a bluetooth connection, because it will request using APIs and DBUS service.

Quoted 5 lines

Can I create a library in the pand directory and reuse the common codes?
I suggest a directive like "--enable-panz" because this library aims provide functions to hide connection details.

I think you get the whole idea of D-Bus wrong and this is not about the current D-Bus support of hcid for the PIN helper.

As I said if you wanna control something from your application via D-Bus then make sure there is a D-Bus enabled pand service is running. If you wanna do it at another level, then I don't see any benefit from it. Why should we create another library only to help your special D-Bus daemon if it is not Bluetooth only. In that case you should simply duplicate the code and integrate it into your daemon directly. There is no need to make things more complicate.

For me the only thing that makes sense if you enhance the pand with D-Bus support and then you use that D-Bus interface in your application.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-01 21:16 UTC #6 lore raw

Hi Marcel,

I suggested provide a DBUS message service in another level because I will try contribute with NetworkManager project:
http://people.redhat.com/dcbw/NetworkManager/index.html

I will analyze these two approachs and define the most interesting solution for my project.

Thanks for your time,
Claudio.

Quoted 41 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 20 lines
My application will provide a DBUS service and APIs to alows user mode applications interact with diferent types of Wireless/Wired connections. This is the reason to create a BlueZ library based on pand and hcitool functions. My application should the able to retrieve bluetooth device information, including establish connection, disconnect, link status, new devices/services, ...
The DBUS service will not be provided at BlueZ level.
The current DBUS message in the hcid is used to handle security/authentication issues. This message has a different proposal. Therefore, use the same macro "--enable-dbus" to enable/disable is not a good ideia.
Another important information is: the user mode application doesn't need root privileges to create a bluetooth connection, because it will request using APIs and DBUS service.
Quoted 5 lines
Can I create a library in the pand directory and reuse the common codes?
I suggest a directive like "--enable-panz" because this library aims provide functions to hide connection details.
=20
I think you get the whole idea of D-Bus wrong and this is not about the current D-Bus support of hcid for the PIN helper.
=20
As I said if you wanna control something from your application via D-Bus then make sure there is a D-Bus enabled pand service is running. If you wanna do it at another level, then I don't see any benefit from it. Why should we create another library only to help your special D-Bus daemon if it is not Bluetooth only. In that case you should simply duplicate the code and integrate it into your daemon directly. There is no need to make things more complicate.
=20
For me the only thing that makes sense if you enhance the pand with D-Bus support and then you use that D-Bus interface in your application.

Quoted 8 lines

=20
Regards
=20
Marcel
=20
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, =

4

Quoted 7 lines

opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-01 23:18 UTC #7 lore raw

Hi Claudio,

Quoted 4 lines

I suggested provide a DBUS message service in another level because I will try contribute with NetworkManager project:
http://people.redhat.com/dcbw/NetworkManager/index.html

then this is not about D-Bus integration. What you need is to make the NetworkManager daemon capable of dealing with Bluetooth PAN connections.

Quoted 2 lines

I will analyze these two approachs and define the most interesting solution for my project.

Both parts are fine with me, but the NetworkManager thing should be done outside of the BlueZ utilities.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-02 15:32 UTC #8 lore raw

Hi Marcel,

After analyze both approachs and consider the future enhancements for BlueZ, implement DBUS service at BlueZ level can be better because it open a new perspective, local services can provided:
- dial-up networking
- business card exchange
- fax
- file transfer
- headset
- audio gateway
- ...

This is very interesting for embedded environment because only the assigned application will handle the request/notifications.

Regarding the service name, I suggest:
"org.bluez.pand"
Is it fine for you?

Do you want a macro to enable this feature at
build time?

Regards,
Claudio.

Quoted 12 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 4 lines
I suggested provide a DBUS message service in another level because I will try contribute with NetworkManager project:
http://people.redhat.com/dcbw/NetworkManager/index.html
=20
then this is not about D-Bus integration. What you need is to make the NetworkManager daemon capable of dealing with Bluetooth PAN connections. =20
Quoted 2 lines
I will analyze these two approachs and define the most interesting solution for my project.

Quoted 11 lines

=20
Both parts are fine with me, but the NetworkManager thing should be done outside of the BlueZ utilities.
=20
Regards
=20
Marcel
=20
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, =

4

Quoted 7 lines

opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
=20

--=20
_______________________________________
Claudio Takahasi
INdT - Nokia Technology Institute
Phone: +55 81 21223034
Recife - Pernambuco - Brazil

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-02 16:24 UTC #9 lore raw

Hi Claudio,

Quoted 11 lines

After analyze both approachs and consider the future enhancements for BlueZ, implement DBUS service at BlueZ level can be better because it open a new perspective, local services can provided:
- dial-up networking
- business card exchange
- fax
- file transfer
- headset
- audio gateway
- ...

Quoted 6 lines

This is very interesting for embedded environment because only the assigned application will handle the request/notifications.
Regarding the service name, I suggest:
"org.bluez.pand"
Is it fine for you?

the name is fine with me.

Quoted 2 lines

Do you want a macro to enable this feature at
build time?

Follow the hcid example. The configure script already detects if D-Bus is available or not.

Make sure that the current tasks (--connect and --listen) of pand are working the same and use --dbus for starting a pand with a D-Bus interface.

Keep all the D-Bus specific things in a separate files.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-09 21:31 UTC #10 lore raw

Hi Marcel,

According with our last discussion. I will to add DBUS service support in the pand. All code related to DBUS will be added=20 in a new file.

The dbus service will be enabled using the argument "--dbus". However if this argument is provide with other arguments like "--connect", "--role" or "--listen" a sharing problem will happen.

What is the behaviour that you want?
1. if "--dbus" was informed no other arguments will be accepted
2. support a shared mode, accepting other arguments.

In this last case, I have to solve a small problem: shared=20 variables, add functions to check the device status, and common codes like this:

create_connection()

static uint16_t role =3D BNEP_SVC_PANU;=09/* Local role (ie service) */ static uint16_t service =3D BNEP_SVC_NAP;=09/* Remote service */

static int  encrypt;
static int  secure;
static int  master;

static char netdev[16] =3D "bnep%d";
static bdaddr_t src_addr =3D *BDADDR_ANY;
static int src_dev =3D -1;

Regards,
Claudio.

Quoted 33 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 11 lines
After analyze both approachs and consider the future enhancements for BlueZ, implement DBUS service at BlueZ level can be better because it open a new perspective, local services can provided:
- dial-up networking
- business card exchange
- fax
- file transfer
- headset
- audio gateway
- ...
Quoted 6 lines
This is very interesting for embedded environment because only the assigned application will handle the request/notifications.
Regarding the service name, I suggest:
"org.bluez.pand"
Is it fine for you?
=20
the name is fine with me.
=20
Quoted 2 lines
Do you want a macro to enable this feature at
build time?
=20
Follow the hcid example. The configure script already detects if D-Bus is available or not.
=20
Make sure that the current tasks (--connect and --listen) of pand are working the same and use --dbus for starting a pand with a D-Bus interface.
=20
Keep all the D-Bus specific things in a separate files.

Quoted 8 lines

=20
Regards
=20
Marcel
=20
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, =

4

Quoted 7 lines

opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
=20

--=20
_______________________________________
Claudio Takahasi
INdT - Nokia Technology Institute
Phone: +55 81 21223034
Recife - Pernambuco - Brazil

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-11 18:27 UTC #11 lore raw

Hi Claudio,

Quoted 11 lines

According with our last discussion. I will to add DBUS service support in the pand. All code related to DBUS will be added in a new file.
The dbus service will be enabled using the argument "--dbus". However if this argument is provide with other arguments like "--connect", "--role" or "--listen" a sharing problem will happen.
What is the behaviour that you want?
1. if "--dbus" was informed no other arguments will be accepted
2. support a shared mode, accepting other arguments.

you should do 1). The rest makes no sense, because if you start it with --dbus you want to control everything via a D-Bus application and pand acts like a proxy.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Peter Wippich

2005-05-12 09:00 UTC #12 lore raw

Hi all,

sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ??? Think I missed something.

Thanks a lot,

Peter

|	Peter Wippich 			Voice: +49 30 46776411		|
|	G&W Instruments GmbH		fax:   +49 30 46776419		|
|	Gustav-Meyer-Allee 25, Geb. 12	Email: [email protected]	|
|	D-13355 Berlin  / Germany					|

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-12 12:41 UTC #13 lore raw

Hi Peter,

Quoted 2 lines

sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ???

look at http://www.freedesktop.org/Software/dbus

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Peter Wippich

2005-05-12 16:21 UTC #14 lore raw

Thank you Marcel !

Ciao,

Peter

Quoted 14 lines · Marcel Holtmann

Hi Peter,
Quoted 2 lines
sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ???
look at http://www.freedesktop.org/Software/dbus
Regards
Marcel

Quoted 10 lines

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

|	Peter Wippich 			Voice: +49 30 46776411		|
|	G&W Instruments GmbH		fax:   +49 30 46776419		|
|	Gustav-Meyer-Allee 25, Geb. 12	Email: [email protected]	|
|	D-13355 Berlin  / Germany					|

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-20 21:04 UTC #15 lore raw

Hi Marcel,

I am sending a initial skeleton of pand DBUS support. I was not able to access BlueZ CVS due firewall permissions.

The path is based on bluez-utils-2.17.

The Makefile.am file was changed. Run aclocal, automake and autoconf tools.

The BLUEZ DBUS PAN service uses the system bus. Check the following items before run the sample client:
1. checking dbus-daemon
type: $ps -ef | grep dbus
"dbus-daemon --system" should be displayed
If not, install dbus-1 package

2. check dbus service configuration file
The file bluez.conf must be located at: /etc/dbus-1/system.d/ See Appendix 1 of DBUS.TXT file located in pand directory
After copy bluez.conf file I suggest restart dbus-daemon

3. running the bluez pan daemon (requires root permission) type $pand --dbus or pand -b --nodetach
check if the pand process is running: $ps -ef | grep pand if not, read the system log: $tail /var/log/syslog

4. extract the client (doesn't require root permission) The client calls basic functions. This is a usage sample for gamming, UPnP, ... A library can be developed to hide D-Bus functions.

Suggestions and comments are welcome!

Regarding dbus-0.33 support, I decided implement using a compatible version(0.23). When a stable version become available we can think on move to the new version.

Claudio.

Quoted 42 lines · Peter Wippich

Thank you Marcel !
Ciao,
Peter
Quoted 14 lines · Marcel Holtmann
Hi Peter,
Quoted 2 lines
sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ???
look at http://www.freedesktop.org/Software/dbus
Regards
Marcel
Quoted 10 lines
-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
|       Peter Wippich                   Voice: +49 30 46776411          |
|       G&W Instruments GmbH            fax:   +49 30 46776419          |
|       Gustav-Meyer-Allee 25, Geb. 12  Email: [email protected]   |
|       D-13355 Berlin  / Germany                                       |

Quoted 10 lines

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

pand/dbus_services.h +206 -0

diff -urNaw bluez-utils-2.17.orig/pand/dbus_services.h bluez-utils-2.17-dbus/pand/dbus_services.h
@@ -0,0 +1,206 @@
+/*
+ *
+ *  BlueZ - Bluetooth protocol stack for Linux
+ *
+ *  Copyright (C) 2005       Claudio Takahasi <[email protected]>
+ *  Copyright (C) 2002-2004  Marcel Holtmann <[email protected]>
+ *
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License version 2 as
+ *  published by the Free Software Foundation;
+ *
+ *  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ *  OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ *  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF THIRD PARTY RIGHTS.
+ *  IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) AND AUTHOR(S) BE LIABLE FOR ANY
+ *  CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES, OR ANY DAMAGES
+ *  WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ *  ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ *  OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ *
+ *  ALL LIABILITY, INCLUDING LIABILITY FOR INFRINGEMENT OF ANY PATENTS,
+ *  COPYRIGHTS, TRADEMARKS OR OTHER RIGHTS, RELATING TO USE OF THIS
+ *  SOFTWARE IS DISCLAIMED.
+ *
+ *
+ *  $Id: pand_dbus_services.h,v 1.12 2005/05/22 13:00:00 holtmann Exp $
+ */
+#ifndef __H_PAND_DBUS_SERVICES_H__ 
+#define __H_PAND_DBUS_SERVICES_H__
+
+#include <stdlib.h>
+#include <stdint.h>
+
+#define ORG_BLUEZ_DBUS_NAME "org.bluez"
+
+#define ORG_BLUEZ_PAND_DBUS_IFACE "org.bluez.pandIface"
+#define ORG_BLUEZ_HCI_DBUS_IFACE  "org.bluez.hciIface"
+
+#define ORG_BLUEZ_PAND_OBJ "/org/bluez/pand"
+#define ORG_BLUEZ_HCI_OBJ  "/org/bluez/hci"
+
+
+// See pand_dbus_conn_req_t, used by actions attribute
+#define PAND_CREATE_BRIDGE_MASK   0x01
+#define PAND_LLA_MASK             0x02
+
+
+// status of request operations
+#define PAND_RESULT_SUCCESS      0x00
+#define PAND_RESULT_FAILED       0x01
+
+// see pand_dev_info_t
+#define PAND_DEV_STATE_READY           0x01
+#define PAND_DEV_STATE_CONNECTED       0x02
+
+
+typedef enum __attribute__((packed))
+{
+	PAND_DBUS_INVALID_TYPE,
+	PAND_DBUS_CHANGE_ROLE_REQ_TYPE,
+	PAND_DBUS_CHANGE_ROLE_REQ_CNF_TYPE,
+	PAND_DBUS_CONN_REQ_TYPE,
+	PAND_DBUS_CONN_REQ_CNF_TYPE,
+	PAND_DBUS_DISCONN_REQ_TYPE,
+	PAND_DBUS_DISCONN_REQ_CNF_TYPE,
+	PAND_DBUS_DISCONN_ALL_REQ_TYPE,
+	PAND_DBUS_DISCONN_ALL_REQ_CNF_TYPE,
+	PAND_DBUS_GET_ACTIVE_CONN_REQ_TYPE,
+	PAND_DBUS_GET_ACTIVE_CONN_REQ_CNF_TYPE,
+	PAND_DBUS_SEARCH_REQ_TYPE,   /* request search for a PAN service */
+	PAND_DBUS_SEARCH_REQ_CNF_TYPE,
+	PAND_DBUS_GET_DEV_INFO_REQ_TYPE,  /* request local bluetooth devices */
+	PAND_DBUS_GET_DEV_INFO_CNF_TYPE,
+	PAND_DBUS_LISTEN_REQ_TYPE,
+	PAND_DBUS_LISTEN_REQ_CNF_TYPE,
+	PAND_DBUS_LISTEN_STOP_REQ_TYPE,
+	PAND_DBUS_LISTEN_STOP_REQ_CNF_TYPE
+}PAND_MESSAGE_TYPE;
+
+
+typedef enum __attribute__((packed))
+{
+	DISABLE_BRIDGE,
+	ENABLE_BRIDGE
+}BRIDGE_MODE;
+
+typedef enum __attribute__((packed))
+{
+	PAND_STATUS_BUSY,
+	PAND_STATUS_CONNECTED,
+	PAND_STATUS_LISTENNING,
+	PAND_STATUS_IDLE
+}PAND_DBUS_STATUS;
+
+
+typedef struct __attribute__((packed))
+{
+	PAND_MESSAGE_TYPE   type;
+	void     *data;
+}pand_dbus_message_format_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t dev_id; /* for address more than one device */
+	uint16_t role;
+	uint16_t service;
+	char     actions;
+	char     bdaddr[18];
+}pand_dbus_conn_req_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t dev_id; /* for address more than one device */
+	char     bdaddr[18];
+}pand_dbus_disconn_req_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t dev_id; /* for address more than one device */
+}pand_dbus_disconn_all_req_t,pand_dbus_get_active_conn_req_t;
+
+/* Generic confirmation result */
+typedef struct __attribute__((packed))
+{
+	PAND_MESSAGE_TYPE     type;
+	char     	      result;
+	char                  data[0];
+}pand_dbus_req_cnf_t;
+
+
+typedef struct __attribute__((packed))
+{
+	uint16_t	dev_id; /* for address more than one device */
+	uint16_t	role;
+	BRIDGE_MODE     bridge_mode;
+	char		master;
+	char		secure;
+	char		encrypt;
+	char		ipv4[16];
+	char		ipv4subnet[16]; 
+}pand_dbus_listen_req_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t	dev_id; /* for address more than one device */
+}pand_dbus_listen_stop_req_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t	dev_id; /* for address more than one device */
+	uint16_t	role;   /* new role*/
+	char     	bdaddr[18];
+}pand_dbus_change_role_req_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t dev_id; /* for address more than one device */
+	uint16_t service;
+}pand_dbus_search_req_t;
+
+typedef struct __attribute__((packed))
+{
+	char status; /* 0-success 1-failed */
+	char *ip; /* ip address if lla was requested */
+}pand_dbus_conn_req_cnf_t;
+
+typedef struct __attribute__((packed))
+{
+	char status;
+	/* device list ???*/	
+}pand_dbus_search_cnf_t;
+
+typedef struct __attribute__((packed))
+{
+	uint16_t dev_id;
+	char     name[8];
+	char     bdaddr[18];
+	uint16_t state; /* see PAND_DEV_STATE_XXXX    TODO  */
+}pand_dev_info_t;
+
+/* Addressing more than one bluetooth device */
+typedef struct __attribute__((packed))
+{
+	//TODO add support for more than one device
+	char num_dev;
+	pand_dev_info_t devices[0];
+}pand_dbus_get_dev_cnf_t;
+
+typedef struct __attribute__((packed))
+{
+	char bdaddr[18];
+	char link_type;
+	char handle;
+	char state;
+	char master;
+}pand_active_conn_t;
+typedef struct __attribute__((packed))
+{
+	//TODO add support for more than one device
+	char num_conn;
+	pand_active_conn_t conn[0];
+}pand_dbus_get_active_conn_cnf_t;
+
+
+#endif /* __H_PAND_DBUS_SERVICES_H__*/

pand/DBUS.TXT +65 -0

65 lines not on this page · raw message

pand/main.c +28 -4

diff -urNaw bluez-utils-2.17.orig/pand/main.c bluez-utils-2.17-dbus/pand/main.c
@@ -54,6 +54,10 @@
 
 #include "pand.h"
 
+#ifdef ENABLE_DBUS
+#include "pand_dbus.h"
+#endif
+
 static uint16_t role = BNEP_SVC_PANU;	/* Local role (ie service) */
 static uint16_t service = BNEP_SVC_NAP;	/* Remote service */

@@ -88,7 +92,10 @@
 	SHOW,
 	LISTEN,
 	CONNECT,
-	KILL
+	KILL,
+#ifdef ENABLE_DBUS
+	DBUS,
+#endif
 } modes;
 
 static void run_devup(char *dev, char *dst, int sk, int nsk)

@@ -417,6 +424,9 @@
 void sig_term(int sig)
 {
 	terminate = 1;
+#ifdef ENABLE_DBUS
+	pand_dbus_stop();
+#endif
 }
 
 int write_pidfile(void)

@@ -488,6 +498,9 @@
 	{ "help",     0, 0, 'h' },
 	{ "listen",   0, 0, 's' },
 	{ "connect",  1, 0, 'c' },
+#ifdef ENABLE_DBUS
+	{ "dbus",     0, 0, 'b' },
+#endif
 	{ "search",   2, 0, 'Q' },
 	{ "kill",     1, 0, 'k' },
 	{ "killall",  0, 0, 'K' },

@@ -510,7 +523,7 @@
 	{ 0, 0, 0, 0 }
 };
 
-static char main_sopts[] = "hsc:k:Kr:e:i:lnp::DQ::AESMC::P:z";
+static char main_sopts[] = "hsc:bk:Kr:e:i:lnp::DQ::AESMC::P:z";
 
 static char main_help[] = 
 	"Bluetooth PAN daemon version " VERSION " \n"

@@ -520,6 +533,9 @@
 	"\t--show --list -l          Show active PAN connections\n"
 	"\t--listen -s               Listen for PAN connections\n"
 	"\t--connect -c <bdaddr>     Create PAN connection\n"
+#ifdef ENABLE_DBUS
+	"\t--dbus    -b              Enable dbus service\n"
+#endif
 	"\t--autozap -z              Disconnect automatically on exit\n"
 	"\t--search -Q[duration]     Search and connect\n"
 	"\t--kill -k <bdaddr>        Kill PAN connection\n"

@@ -560,7 +576,11 @@
 			mode = CONNECT;
 			dst  = strdup(optarg);
 			break;
-
+#ifdef ENABLE_DBUS
+		case 'b':
+			mode = DBUS;
+			break;
+#endif
 		case 'Q':
 			mode = CONNECT;
 			if (optarg)

@@ -729,8 +749,12 @@
 	case LISTEN:
 		do_listen();
 		break;
+#ifdef ENABLE_DBUS
+	case DBUS:
+		pand_dbus_start();
+		break;
+#endif
 	}
-
 	if (pidfile)
 		unlink(pidfile);

pand/Makefile.am +22 -3

diff -urNaw bluez-utils-2.17.orig/pand/Makefile.am bluez-utils-2.17-dbus/pand/Makefile.am
@@ -2,12 +2,31 @@
 #  $Id: Makefile.am,v 1.5 2005/04/21 21:34:10 holtmann Exp $
 #
 
+if DBUS
+    pand_dbus_inc_hdr = dbus_services.h
+    pand_dbus_sources = pand_dbus.c pand_dbus_util.c pand_dbus_commands.c
+    pand_dbus_libs    = @DBUS_LIBS@ -ldbus-glib-1 -lbridge -lpthread
+    pand_dbus_cflags  = -DENABLE_DBUS -DDBUS_API_SUBJECT_TO_CHANGE
+    pand_dbus_inc     = -I/usr/include/dbus-1.0 -I/usr/lib/dbus-1.0/include -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/local/include
+else
+    pand_dbus_inc_hdr =
+    pand_dbus_sources =
+    pand_dbus_libs    =
+    pand_dbus_cflags  =
+    pand_dbus_inc     =
+endif
+
+
+
 bin_PROGRAMS = pand
 
-pand_SOURCES = main.c pand.h bnep.c sdp.c
-pand_LDADD = @BLUEZ_LIBS@
+pand_SOURCES = main.c pand.h bnep.c sdp.c $(pand_dbus_sources)
+pand_LDADD = @BLUEZ_LIBS@ $(pand_dbus_libs)
+INCLUDES = $(pand_dbus_inc)
+includedir = @includedir@/bluetooth
+include_HEADERS = $(pand_dbus_inc_hdr)
 
-AM_CFLAGS = @BLUEZ_CFLAGS@
+AM_CFLAGS = @BLUEZ_CFLAGS@ $(pand_dbus_cflags)
 
 man_MANS = pand.1

pand/pand_dbus.c +721 -0

721 lines not on this page · raw message

pand/pand_dbus_commands.c +514 -0

514 lines not on this page · raw message

pand/pand_dbus_commands.h +45 -0

45 lines not on this page · raw message

pand/pand_dbus.h +41 -0

41 lines not on this page · raw message

pand/pand_dbus_util.c +102 -0

102 lines not on this page · raw message

pand/pand_dbus_util.h +55 -0

55 lines not on this page · raw message

Marcel Holtmann

maintainer

2005-05-23 10:27 UTC #16 lore raw

Hi Claudio,

Quoted 4 lines

I am sending a initial skeleton of pand DBUS support. I was not able to access BlueZ CVS due firewall permissions.
The path is based on bluez-utils-2.17.

no problem the 2.17 is quite up-to-date.

Quoted 1 line

Suggestions and comments are welcome!

please use only the file dbus.c and put needed function prototypes into pand.h. The dbus_services.h is not needed. Put the declarations into dbus.c.

Don't provide and "hcid" service within pand. If you wanna provide HCI task then hide them inside the pand D-Bus. Use org.bluez.pand as base name for all stuff.

Never use dev_id. You can use the source Bluetooth address instead and then convert it to dev_id internally. Don't use bdaddr_t at all. Simply use the string representation of the Bluetooth device address.

Don't use char to represent an one byte value. Use uint8_t.

Once if we agree on the changes to Makefile.am, pand.h and main.c, then I can commit a simple framework with an empty dbus.c and you only need to send me changes to dbus.c from then on.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-23 15:02 UTC #17 lore raw

Hi Marcel,

I will fix your suggestions.

Regarding our suggestion about the service name. I consider the following basic features essential to establish and control connections:
* Search for a service (provided by sdptool)
* Display local devices (provided by hcitool)
* Switch master/slave role (provided by hcitool)
* Listen for PAN connections (provided by pand)
* Connect (provided by pand)
* Disconnect (provided by pand)
* Display active connections (provided by pand)

According to D-BUS specification when messages are received over a D-BUS connection,
they are sent to a specific object, not to the application as a whole. Each object supports
one or more interfaces.
=20
We have two options:
1. define one object with only one interface

    eg:   =20
        object:org.bluez.pand
        interface:org.bluez.pand
2. define one object with multiple interfaces
    eg:   =20
        object:org.bluez.pand
        interface: org.bluez.pand
        interface: org.bluez.sdp
        interface: org.bluez.hci

For the both cases, pand will be the daemon responsible for provide this interfaces/services.
What approach do you preffer?
The last alternative is better because it is possible define more detailed rules in the dbus configuration file.=20

Another point is the header for dbus services, the client application must know the message format, therefore the header file(dbus_services.h) will be required. In the future, develop a library hidding D-BUS message details is a better option.

Regards,
Claudio.

Quoted 37 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 4 lines
I am sending a initial skeleton of pand DBUS support. I was not able to access BlueZ CVS due firewall permissions.
The path is based on bluez-utils-2.17.
=20
no problem the 2.17 is quite up-to-date.
=20
Quoted 1 line
Suggestions and comments are welcome!
=20
please use only the file dbus.c and put needed function prototypes into pand.h. The dbus_services.h is not needed. Put the declarations into dbus.c.
=20
Don't provide and "hcid" service within pand. If you wanna provide HCI task then hide them inside the pand D-Bus. Use org.bluez.pand as base name for all stuff.
=20
Never use dev_id. You can use the source Bluetooth address instead and then convert it to dev_id internally. Don't use bdaddr_t at all. Simply use the string representation of the Bluetooth device address. =20
Don't use char to represent an one byte value. Use uint8_t. =20
Once if we agree on the changes to Makefile.am, pand.h and main.c, then I can commit a simple framework with an empty dbus.c and you only need to send me changes to dbus.c from then on.
=20
Regards
=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------

Quoted 9 lines

This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=3D7412&alloc_id=3D16344&op=3Dclick
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-24 09:07 UTC #18 lore raw

Hi Claudio,

Quoted 28 lines

Regarding our suggestion about the service name. I consider the following basic features essential to establish and control connections:
* Search for a service (provided by sdptool)
* Display local devices (provided by hcitool)
* Switch master/slave role (provided by hcitool)
* Listen for PAN connections (provided by pand)
* Connect (provided by pand)
* Disconnect (provided by pand)
* Display active connections (provided by pand)
According to D-BUS specification when messages are received over a D-BUS connection,
they are sent to a specific object, not to the application as a whole. Each object supports
one or more interfaces.
We have two options:
1. define one object with only one interface
    eg:    
        object:org.bluez.pand
        interface:org.bluez.pand
2. define one object with multiple interfaces
    eg:    
        object:org.bluez.pand
        interface: org.bluez.pand
        interface: org.bluez.sdp
        interface: org.bluez.hci

Quoted 5 lines

For the both cases, pand will be the daemon responsible for provide this interfaces/services.
What approach do you preffer?
The last alternative is better because it is possible define more detailed rules in the dbus configuration file.

this might be, but until we don't fully understand what functionality we need to provide over D-Bus I like to go with 1). In general the first version of an API sucks and I don't wanna spread it over different interfaces and have to deprecate them.

Quoted 4 lines

Another point is the header for dbus services, the client application must know the message format, therefore the header file(dbus_services.h) will be required. In the future, develop a library hidding D-BUS message details is a better option.

Then make it dbus.h for the service definitions and move the rest into dbus.c and the function declaration into pand.h. For now copy the dbus.h to the client application and we will sort out the details of it later.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-25 22:20 UTC #19 lore raw

Hi Marcel,

A new patch is available in the link below:
http://www.indt.org.br/maemo/bluez/

The DBUS.TXT file contains information to run the sample and how configure the service.

Suggestions and comments are welcome!

Regards,
Claudio.

Quoted 4 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 2 lines
Regarding our suggestion about the service name. I consider the following basic features essential to establish and control connections=

:

Quoted 31 lines

Quoted 26 lines
* Search for a service (provided by sdptool)
* Display local devices (provided by hcitool)
* Switch master/slave role (provided by hcitool)
* Listen for PAN connections (provided by pand)
* Connect (provided by pand)
* Disconnect (provided by pand)
* Display active connections (provided by pand)
According to D-BUS specification when messages are received over a D-BUS connection,
they are sent to a specific object, not to the application as a whole. Each object supports
one or more interfaces.
We have two options:
1. define one object with only one interface
    eg:
        object:org.bluez.pand
        interface:org.bluez.pand
2. define one object with multiple interfaces
    eg:
        object:org.bluez.pand
        interface: org.bluez.pand
        interface: org.bluez.sdp
        interface: org.bluez.hci
Quoted 5 lines
For the both cases, pand will be the daemon responsible for provide this interfaces/services.
What approach do you preffer?
The last alternative is better because it is possible define more detailed rules in the dbus configuration file.

Quoted 8 lines

=20
this might be, but until we don't fully understand what functionality we need to provide over D-Bus I like to go with 1). In general the first version of an API sucks and I don't wanna spread it over different interfaces and have to deprecate them.
=20
Quoted 2 lines
Another point is the header for dbus services, the client application must know the message format, therefore the header file(dbus_services.h=

)

Quoted 1 line

Quoted 1 line
will be required. In the future, develop a library hidding D-BUS messag=

e

Quoted 14 lines

Quoted 1 line
details is a better option.
=20
Then make it dbus.h for the service definitions and move the rest into dbus.c and the function declaration into pand.h. For now copy the dbus.h to the client application and we will sort out the details of it later. =20
Regards
=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------

Quoted 9 lines

This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=3D7412&alloc_id=3D16344&op=3Dclick
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate online with coworkers and clients while avoiding the high cost of travel and communications. There is no equipment to buy and you can meet as often as you want. Try it free.http://ads.osdn.com/?ad_id=7402&alloc_id=16135&op=click
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-26 09:08 UTC #20 lore raw

Hi Claudio,

Quoted 2 lines

A new patch is available in the link below:
http://www.indt.org.br/maemo/bluez/

lets welcome the Nokia 770 development ;)

Quoted 2 lines

The DBUS.TXT file contains information to run the sample and how configure the service.

We must make sure that we install a default service file, but we can sort this out later.

Quoted 1 line

Suggestions and comments are welcome!

We must really talk about the coding style at some point, but this is not important at the moment. We follow the kernel coding style and I wanna keep it that way.

While looking through the whole source I realized that you used Glib and Pthread. This introduces a dependency chain that I don't wanna take. We must be really careful about this, because not all embedded systems can deal with it. We had a hard time to remove the Glib dependency from hcid and I am not going to introduce it again. For Pthread you might convince me, but there must be really good arguments.

What is libbridge? I don't found it on any of my systems.

Regards

Marcel

-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate online with coworkers and clients while avoiding the high cost of travel and communications. There is no equipment to buy and you can meet as often as you want. Try it free.http://ads.osdn.com/?ad_id=7402&alloc_id=16135&op=click
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-26 21:22 UTC #21 lore raw

Hi Marcel,

1. Glib
I will fix it. I found another approach using dbus main loop function.

2. libbridge
It should be removed. It is not being used at the momment. I will fix the Makefile.am
But in the the future, I would suggest provide automatic bridge creation. Using libbridge it's possible create bridge and add bnep interfaces.

3. Coding style
sorry, I try do my best. I am not familiar with linux kernel coding style.

4. PThread=20
I will remove Posix Thread code.

Thanks for our comments!
I will send another patch until next monday.

Regards,
Claudio.

Quoted 29 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 2 lines
A new patch is available in the link below:
http://www.indt.org.br/maemo/bluez/
=20
lets welcome the Nokia 770 development ;)
=20
Quoted 2 lines
The DBUS.TXT file contains information to run the sample and how configure the service.
=20
We must make sure that we install a default service file, but we can sort this out later.
=20
Quoted 1 line
Suggestions and comments are welcome!
=20
We must really talk about the coding style at some point, but this is not important at the moment. We follow the kernel coding style and I wanna keep it that way.
=20
While looking through the whole source I realized that you used Glib and Pthread. This introduces a dependency chain that I don't wanna take. We must be really careful about this, because not all embedded systems can deal with it. We had a hard time to remove the Glib dependency from hcid and I am not going to introduce it again. For Pthread you might convince me, but there must be really good arguments.
=20
What is libbridge? I don't found it on any of my systems. =20
Regards

Quoted 8 lines

=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborat=

e

Quoted 1 line

online with coworkers and clients while avoiding the high cost of travel =

and

Quoted 2 lines

communications. There is no equipment to buy and you can meet as often as you want. Try it free.http://ads.osdn.com/?ad_id=3D7402&alloc_id=3D16135&=

op=3Dclick

Quoted 5 lines

_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-26 21:40 UTC #22 lore raw

Hi Claudio,

Quoted 2 lines

1. Glib
I will fix it. I found another approach using dbus main loop function.

perfect.

Quoted 5 lines

2. libbridge
It should be removed. It is not being used at the momment. I will fix the Makefile.am
But in the the future, I would suggest provide automatic bridge creation. Using libbridge it's possible create bridge and add bnep interfaces.

I am not against it, but I think we need to make it optional. Does it make sense to put that into pand or leave it in the client?

Quoted 2 lines

3. Coding style
sorry, I try do my best. I am not familiar with linux kernel coding style.

Maybe you wanna read the paper from Greg KH on the BlueZ page. We will sort that out later.

Quoted 2 lines

4. PThread
I will remove Posix Thread code.

We don't really need threads.

Quoted 1 line

I will send another patch until next monday.

Keep up the good work.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-30 22:42 UTC #23 lore raw

Hi Marcel,

I am sending a new patch to be evaluated. Regarding glib dependency we need discuss this subject again. If you see the hcid code, there is a file (glib-ectomy.c) that implements the event processing, this code is a partial copy of glib.

What do you prefer? Repeat the glib-ectomy.c code
or use glib?

For Maemo, glib is not problem. It is included in
the SDK.

Regarding pthread I put two versions to be evaluated:
1. using pthread
In my opinion because it save memory.
http://www.indt.org.br/maemo/bluez/bluez-dbus-thread.diff
=20
2. using fork
In my opinion this is not a feasible approach because when listening is active two process=20 will be running(a listening and another processing D-Bus messages).
http://www.indt.org.br/maemo/bluez/bluez-dbus-fork.diff

Regards,
Claudio.

Quoted 11 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 2 lines
1. Glib
I will fix it. I found another approach using dbus main loop function.
=20
perfect.
=20
Quoted 4 lines
2. libbridge
It should be removed. It is not being used at the momment. I will fix the Makefile.am
But in the the future, I would suggest provide automatic bridge creatio=

n.

Quoted 7 lines

Quoted 1 line
Using libbridge it's possible create bridge and add bnep interfaces.
=20
I am not against it, but I think we need to make it optional. Does it make sense to put that into pand or leave it in the client? =20
Quoted 2 lines
3. Coding style
sorry, I try do my best. I am not familiar with linux kernel coding sty=

le.

Quoted 10 lines

=20
Maybe you wanna read the paper from Greg KH on the BlueZ page. We will sort that out later.
=20
Quoted 2 lines
4. PThread
I will remove Posix Thread code.
=20
We don't really need threads.
=20
Quoted 1 line
I will send another patch until next monday.

Quoted 15 lines

=20
Keep up the good work.
=20
Regards
=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=3Doffad-ysdn-ostg-q22=

005

Quoted 5 lines

_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-05-30 23:05 UTC #24 lore raw

Hi Claudio,

Quoted 8 lines

I am sending a new patch to be evaluated. Regarding glib dependency we need discuss this subject again. If you see the hcid code, there is a file (glib-ectomy.c) that implements the event processing, this code is a partial copy of glib.
What do you prefer? Repeat the glib-ectomy.c code or use glib?

the glib-ectomy is only still there, because I was too lazy to replace it with a clean main loop using poll() implementation.

Quoted 2 lines

For Maemo, glib is not problem. It is included in the SDK.

This is not a valid argument. Creating a massive dependency chain is always bad. Using Glib inside bluez-utils is no option.

Quoted 11 lines

Regarding pthread I put two versions to be evaluated:
1. using pthread
In my opinion because it save memory.
http://www.indt.org.br/maemo/bluez/bluez-dbus-thread.diff
2. using fork
In my opinion this is not a feasible approach because when listening is active two process will be running(a listening and another processing D-Bus messages).
http://www.indt.org.br/maemo/bluez/bluez-dbus-fork.diff

Even without looking at the patches, I don't like any of it. I like to see a nice main loop and then using poll(). We don't need threads or forking out processes.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-06-03 20:34 UTC #25 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Marcel,

A new patch is available using a main loop and poll to wait for some event= =20
on a file descriptor list (D-Bus and Bluetooth).

All Glib and pthread dependencies were remove.

http://www.indt.org.br/maemo/bluez/bluez-dbus-poll.diff

Regards,
Claudio.

Quoted 40 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 8 lines
I am sending a new patch to be evaluated. Regarding glib dependency we need discuss this subject again. If you see the hcid code, there is a file (glib-ectomy.c) that implements the event processing, this code is a partial copy of glib.
What do you prefer? Repeat the glib-ectomy.c code or use glib?
=20
the glib-ectomy is only still there, because I was too lazy to replace it with a clean main loop using poll() implementation.
=20
Quoted 2 lines
For Maemo, glib is not problem. It is included in the SDK.
=20
This is not a valid argument. Creating a massive dependency chain is always bad. Using Glib inside bluez-utils is no option. =20
Quoted 11 lines
Regarding pthread I put two versions to be evaluated:
1. using pthread
In my opinion because it save memory.
http://www.indt.org.br/maemo/bluez/bluez-dbus-thread.diff
2. using fork
In my opinion this is not a feasible approach because when listening is active two process will be running(a listening and another processing D-Bus messages).
http://www.indt.org.br/maemo/bluez/bluez-dbus-fork.diff
=20
Even without looking at the patches, I don't like any of it. I like to see a nice main loop and then using poll(). We don't need threads or forking out processes.
=20
Regards
=20
Marcel
=20

Quoted 8 lines

=20
=20
=20
-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=3Doffad-ysdn-ostg-q22=

005

Quoted 5 lines

_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-06-03 20:38 UTC #26 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Claudio,

Quoted 1 line

http://www.indt.org.br/maemo/bluez/bluez-dbus-poll.diff

this gives me a 404 error.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-06-03 21:31 UTC #27 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Sorry,

Now, It is available.

Regards,
Claudio.

Quoted 15 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 1 line
http://www.indt.org.br/maemo/bluez/bluez-dbus-poll.diff
=20
this gives me a 404 error.
=20
Regards
=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you sho=

tput

Quoted 1 line

a projector? How fast can you ride your desk chair down the office luge t=

rack?

Quoted 7 lines

If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-06-03 22:10 UTC #28 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

In reply to Claudio Takahasi

Hi Claudio,

Quoted 4 lines

A new patch is available using a main loop and poll to wait for some event on a file descriptor list (D-Bus and Bluetooth).
All Glib and pthread dependencies were remove.

I just had some some time for a quick review. It is not fully what I had in mind, but it is something from what we can work onwards. I think it is a good idea to include it into release 2.18 and then start on improving it.

Since I am away the next week you might wanna address the coding style issues in that time. Read the paper from Greg KH from the OLS 2002. I think it gives a very good understanding on it.

Regards

Marcel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-06-06 13:18 UTC #29 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Marcel,

Probably you are not satisfied with D-Bus watches. I tryed remove=20 it, but it is not possible.=20
According with D-Bus documentation is not allowed read/write directly in the file descriptor. DBusWatch must be used for main loop integration.

Another important approach is develop a library for D-Bus function hidding. The sample client is just a usage sample and a better idea is use a common library for all applications: gamming, UPnP, messenger, ...

Regarding the coding style I will review it.

Regards,
Claudio.

Quoted 3 lines · Marcel Holtmann

Hi Claudio,
=20
Quoted 1 line
A new patch is available using a main loop and poll to wait for some ev=

ent

Quoted 14 lines

Quoted 3 lines
on a file descriptor list (D-Bus and Bluetooth).
All Glib and pthread dependencies were remove.
=20
I just had some some time for a quick review. It is not fully what I had in mind, but it is something from what we can work onwards. I think it is a good idea to include it into release 2.18 and then start on improving it.
=20
Since I am away the next week you might wanna address the coding style issues in that time. Read the paper from Greg KH from the OLS 2002. I think it gives a very good understanding on it.
=20
Regards

Quoted 8 lines

=20
Marcel
=20
=20
=20
=20
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you sho=

tput

Quoted 1 line

a projector? How fast can you ride your desk chair down the office luge t=

rack?

Quoted 7 lines

If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-06-14 11:37 UTC #30 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Marcel,

There is a new patch available.

http://www.indt.org.br/maemo/bluez/bluez-dbus-poll-v00.02.diff

I read the Greg K.H. paper, I tryed adjust the code following his suggestion. Probably there are points that remain incorrect. Send me suggestions and comments that I will fix it.

I am working on library for hidding D-BUS details. I don' t=20 know how provide it to the developers. Basically, this library will hide the code inside the pand_dbus_client.tar.gz

Another subject that I want introduce is bridge creation. I think we discuss this subject before. Two approachs can be applied:
1. put in the pand daemon under optional feature
advantages: can be shared between dbus and no-dbus=20
disadvantages: libbridge dependency=20
2. put it in a library(commented previously)
advantages: This feature can be added with another=20
features/functions, like D-Bus message=20 functions, Zeroconf integration (UDHCP) or HOWL. UDHCP and HOWL D-BUS services are being developed.
disadvantages: no-dbus pand daemon will not be able to=20
control bridges

Regards,
Claudio.

Quoted 10 lines · Claudio Takahasi

Hi Marcel,
=20
Probably you are not satisfied with D-Bus watches. I tryed remove it, but it is not possible.
According with D-Bus documentation is not allowed read/write directly in the file descriptor. DBusWatch must be used for main loop integration. =20
Another important approach is develop a library for D-Bus function hidding. The sample client is just a usage sample and a better idea is use a common library for all applications: gamming, UPnP, messenger, .=

..

Quoted 10 lines

=20
Regarding the coding style I will review it.
=20
Regards,
Claudio.
=20
Quoted 3 lines · Marcel Holtmann
Hi Claudio,
Quoted 1 line
A new patch is available using a main loop and poll to wait for some =

event

Quoted 5 lines

Quoted 5 lines
Quoted 3 lines
on a file descriptor list (D-Bus and Bluetooth).
All Glib and pthread dependencies were remove.
I just had some some time for a quick review. It is not fully what I ha=

d

Quoted 10 lines

Quoted 10 lines
in mind, but it is something from what we can work onwards. I think it is a good idea to include it into release 2.18 and then start on improving it.
Since I am away the next week you might wanna address the coding style issues in that time. Read the paper from Greg KH from the OLS 2002. I think it gives a very good understanding on it.
Regards

Quoted 7 lines

Quoted 7 lines
Marcel
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you s=

hotput

Quoted 1 line

Quoted 1 line
a projector? How fast can you ride your desk chair down the office luge=

track?

Quoted 8 lines

Quoted 7 lines
If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=3D20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-06-17 21:59 UTC #31 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Marcel,

There is a new patch available.

http://www.indt.org.br/maemo/zeroconf/bluez/bluez-dbus-poll-v00.03.diff

Changes from last release:
- comments fixed
- message handle optimized
- created a library for clients development

Now, the test client is based on a library.
http://www.indt.org.br/maemo/zeroconf/bluez/libbluezdbus-0.1.tar.gz
The sample directory contains the test application:
"client" can be used to request services "sig_listener" is used to listen to signals: new connections,=20 disconnection, role changed, ..

Currently, the listener is not integrated with the requester because Glib main loop integration is required. I will develop a better client for demonstration purpose as son as possible.

Bugs:
- start service is not working

TODO:
-organize the interfaces names/services better.=20 sdptool, pand, dund features should be provided using D-Bus. Align with Maemo gwconnect??? -Develop a new version supporting the new D-Bus version

QUESTIONS:
-Where put library creation?
1. put in the pand daemon under optional feature
advantages: can be shared between dbus and no-dbus.
The control is easier.
disadvantages: libbridge dependency
2. put it in a library(commented previously)
advantages: This feature can be added with another
features/functions, like D-Bus message functions, Zeroconf integration (UDHCP) or HOWL. UDHCP and HOWL D-BUS services are being developed.
disadvantages: no-dbus pand daemon will not be able to
control bridges and all clients will be allowed to control bridges.

The library is not complete. I am sending to see the community feedback.

Regards,
Claudio.

Quoted 26 lines · Claudio Takahasi

Hi Marcel,
=20
There is a new patch available.
=20
http://www.indt.org.br/maemo/bluez/bluez-dbus-poll-v00.02.diff
=20
I read the Greg K.H. paper, I tryed adjust the code following his suggestion. Probably there are points that remain incorrect. Send me suggestions and comments that I will fix it. =20
I am working on library for hidding D-BUS details. I don' t know how provide it to the developers. Basically, this library will hide the code inside the pand_dbus_client.tar.gz =20
Another subject that I want introduce is bridge creation. I think we discuss this subject before. Two approachs can be applied:
1. put in the pand daemon under optional feature
advantages: can be shared between dbus and no-dbus
disadvantages: libbridge dependency
2. put it in a library(commented previously)
advantages: This feature can be added with another
features/functions, like D-Bus message functions, Zeroconf integration (UDHCP) or HOWL. UDHCP and HOWL D-BUS services are being developed.
disadvantages: no-dbus pand daemon will not be able to
control bridges

Quoted 14 lines

=20
=20
=20
Regards,
Claudio.
=20
=20
Quoted 6 lines · Claudio Takahasi
Hi Marcel,
Probably you are not satisfied with D-Bus watches. I tryed remove it, but it is not possible.
According with D-Bus documentation is not allowed read/write directly in the file descriptor. DBusWatch must be used for main loop integratio=

n.

Quoted 4 lines

Quoted 4 lines
Another important approach is develop a library for D-Bus function hidding. The sample client is just a usage sample and a better idea is use a common library for all applications: gamming, UPnP, messenger,=

...

Quoted 10 lines

Quoted 10 lines
Regarding the coding style I will review it.
Regards,
Claudio.
Quoted 3 lines · Marcel Holtmann
Hi Claudio,
Quoted 1 line
A new patch is available using a main loop and poll to wait for som=

e event

Quoted 5 lines

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines
on a file descriptor list (D-Bus and Bluetooth).
All Glib and pthread dependencies were remove.
I just had some some time for a quick review. It is not fully what I =

had

Quoted 1 line

Quoted 1 line
Quoted 1 line
in mind, but it is something from what we can work onwards. I think i=

t

Quoted 4 lines

Quoted 4 lines
Quoted 4 lines
is a good idea to include it into release 2.18 and then start on improving it.
Since I am away the next week you might wanna address the coding styl=

e

Quoted 5 lines

Quoted 5 lines
Quoted 5 lines
issues in that time. Read the paper from Greg KH from the OLS 2002. I think it gives a very good understanding on it.
Regards

Quoted 7 lines

Quoted 7 lines
Quoted 7 lines
Marcel
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you=

shotput

Quoted 1 line

Quoted 1 line
Quoted 1 line
a projector? How fast can you ride your desk chair down the office lu=

ge track?

Quoted 2 lines

Quoted 2 lines
Quoted 2 lines
If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=3D2=

0

Quoted 7 lines

Quoted 6 lines
Quoted 5 lines
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Marcel Holtmann

maintainer

2005-06-18 23:28 UTC #32 lore raw

Re: [Bluez-devel] DBUS service support - poll patch

Hi Claudio,

Quoted 8 lines

There is a new patch available.
http://www.indt.org.br/maemo/zeroconf/bluez/bluez-dbus-poll-v00.03.diff
Changes from last release:
- comments fixed
- message handle optimized
- created a library for clients development

thanks for the new patch. I didn't had time to review it, but I think we really get near what we need. Maybe I have to do another release before I can include your patch.

Regards

Marcel

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Claudio Takahasi

author

2005-05-12 12:45 UTC #33 lore raw

In reply to Peter Wippich

Hi Peter,

You can find more information on the references below:

http://www.freedesktop.org/Software/dbus
http://www-106.ibm.com/developerworks/linux/library/l-dbus.html

Basically, DBUS is system for interprocess communication.

Where/Why PAND DBUS service can be usefull:
- pand required root permission
- user applications are not allowed to establish connection or control the BT device/connections
- For embedded devices(cell phones, PDAs,..), applications like gamming, UPnP, etc.
want to control BT connections, search services, ... and there isn't API for address this.

Obviously, there is a hard work to do. Bridge creation/control, IP assignme= nt,=20
service name resolution have to be improved too in order to provide seamles= s=20
connections.

Regards,
Claudio.

Quoted 16 lines · Peter Wippich

=20
Hi all,
=20
sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ??? Think I missed something.
=20
Thanks a lot,
=20
Peter
=20
=20
|       Peter Wippich                   Voice: +49 30 46776411          |
|       G&W Instruments GmbH            fax:   +49 30 46776419          |
|       Gustav-Meyer-Allee 25, Geb. 12  Email: [email protected]   |
|       D-13355 Berlin  / Germany                                       |

Quoted 12 lines

=20
=20
-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=3D7393&alloc_id=3D16281&op=3Dclick
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Peter Wippich

2005-05-12 16:23 UTC #34 lore raw

Thank you Claudio !

May be this will be interesting for a project we are currently working on. I'll have a look at it.

Ciao,
Peter

Quoted 19 lines · Claudio Takahasi

Hi Peter,
You can find more information on the references below:
http://www.freedesktop.org/Software/dbus
http://www-106.ibm.com/developerworks/linux/library/l-dbus.html
Basically, DBUS is system for interprocess communication.
Where/Why PAND DBUS service can be usefull:
- pand required root permission
- user applications are not allowed to establish connection or control the BT device/connections
- For embedded devices(cell phones, PDAs,..), applications like gamming, UPnP, etc.
want to control BT connections, search services, ... and there isn't API for address this.

Quoted 1 line

Obviously, there is a hard work to do. Bridge creation/control, IP assi=

gnment,

Quoted 1 line

service name resolution have to be improved too in order to provide sea=

mless

Quoted 21 lines

connections.
Regards,
Claudio.
Quoted 13 lines · Peter Wippich
Hi all,
sorry for asking a possible dump question, but:
can anybody point me to a source what this D-BUS stuff is ??? Think I missed something.
Thanks a lot,
Peter
| Peter Wippich Voice: +49 30 46776411 =

|

Quoted 1 line

Quoted 1 line
| G&W Instruments GmbH fax: +49 30 46776419 =

|

Quoted 1 line

Quoted 1 line
| Gustav-Meyer-Allee 25, Geb. 12 Email: [email protected]=

|

Quoted 1 line

Quoted 1 line
| D-13355 Berlin / Germany =

|

Quoted 14 lines

Quoted 12 lines
-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=3D7393&alloc_id=3D16281&op=3Dclick
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Quoted 10 lines

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_ids93&alloc_id=16281&op=CCk
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel

|	Peter Wippich 			Voice: +49 30 46776411		|
|	G&W Instruments GmbH		fax:   +49 30 46776419		|
|	Gustav-Meyer-Allee 25, Geb. 12	Email: [email protected]	|
|	D-13355 Berlin  / Germany					|

-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
Bluez-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel